Purpose. This study aims to compare the dosimetric characteristics of 3D-printed bolus materials acrylonitrile butadiene styrene (ABS), polylactic acid (PLA) with recently introduced natural rubber (NR) bolus. Materials and Methods. We employed Monte Carlo simulation to evaluate ABS, PLA, and NR boluses of thicknesses 0.5, 1.0, and 1.5 cm under 6, 9, and 16 MeV electron beam irradiation. Percentage depth dose (PDD) data was analyzed to evaluate dosimetry parameters. Dosimetric stability under varying air-gap conditions was assessed by analyzing PDD curves under air-gap sizes of (0, 1, 3, 5 mm).Results.NR and ABS showed similar dosimetric profiles, whereas PLA showed enhanced deeper-tissue protection and provided 0.7–10.2% higher surface dose (SD). PLA also exhibited the highest stability, with SD and R90 deviations limited to 1% and 1.2%, respectively, under varying air-gap sizes. At 6 and 9 MeV, a 0.5 cm bolus failed to provide a single dose of 90% of maximum dose for nearly all tested materials Conclusion. NR demonstrated comparable dosimetric performance to ABS, serving as a viable substitute. PLA was optimal for maximizing SD and distal tissue sparing while exhibiting the lowest air-gap sensitivity. For tumors located several centimeters deep from the surface, 16 MeV electron beams were suitable. Furthermore, Dual hotspots were identified a substantial risk of localized normal tissue toxicity.

Purpose. This study aims to evaluate how different shield materials and aperture sizes impact dosimetric properties within the planning target volume (PTV) across various electron beam energies. Material and Methods. Monte Carlo simulations were performed using a 30 × 30 × 30 cm³ solid water phantom covered with a custom bolus-shield assembly. The phantom was irradiated with 6, 9, and 16 MeV electron beam energies. Shield materials included polylactic acid (PLA), acrylonitrile butadiene styrene (ABS), and lead alloy, while the bolus was made of ABS. Three configurations were tested, each pairing a fixed 0.5 cm thick ABS bolus with one of the shield materials (thicknesses were adjusted based on material properties). For each configuration, aperture sizes of 2 × 2, 5 × 5, and 8 × 8 cm² were evaluated. Reference simulations were conducted for each energy using a 0.5 cm ABS bolus covering the entire phantom. Results. Our findings demonstrate that although ABS and PLA require thicker layers and result in wider penumbras than other materials, they are promising shield materials for low- and medium-energy electron beams. These materials improved surface dose coverage, offered superior deep-tissue protection, and eliminated dual hot spots yielding more favorable distributions. Moreover, sensitivity tests confirmed the model’s robustness against setup errors up to ± 5°. However, larger angles introduced obliquity effects, establishing a 5° tolerance limit for reliable clinical applications. Conclusion. This study demonstrates that traditional lead shields can be effectively replaced by a custom 3D-printed ABS or PLA unit that functions simultaneously as a bolus and shield. This approach is particularly effective for low and medium electron energies with small-to-moderate apertures, where it enhances surface dose and protects deep tissues more effectively. Ultimately, these findings confirm that tissue-equivalent polymer shields can satisfy clinical radiobiological requirements, offering a viable, non-toxic alternative to lead to optimizing patient safety during superficial cancer treatments.

Stroke remains a leading cause of global mortality and disability, affecting millions of individuals annually. Computational modeling quantifies the complex pathophysiology of an ischemic stroke, paving the way for personalized therapeutic strategies. This review explores the transformative role of computational modeling and simulation in advancing ischemic stroke research and clinical translation. We systematically examine multi-scale mathematical models, including low-dimensional (0D/1D) hemodynamic representations, high-fidelity 3D computational fluid dynamics (CFD), and fluid-structure interaction (FSI) simulations, to elucidate key hemodynamic parameters such as wall shear stress (WSS), the oscillatory shear index (OSI), and the endothelial cell activation potential (ECAP), which are critical in thrombosis, plaque stability, and stroke progression. Furthermore, we highlight the application of these models in optimizing acute ischemic stroke treatments, including intravenous thrombolysis and mechanical thrombectomy, and in pioneering emerging strategies such as in silico trials and milli-spinner thrombectomy. Despite significant progress, challenges remain in standardization, real-time clinical integration, and model validation. Looking forward, we discuss the integration of multi-scale modeling, artificial intelligence, and patient-specific data toward the development of predictive digital twins and personalized therapeutic frameworks. By bridging mechanistic insights with clinical innovations, computational approaches are poised to redefine stroke care, thus enabling more precise, timely, and effective interventions.

Jumbo bacteriophages possess exceptionally large capsids accommodating genomes that encode additional proteins, which support their infection and replication. A distinctive structural element, known as the inner body, has been observed in a number of phiKZ-like phage particles and its proteins are believed to play an essential role in phage genome organization and ejection. However, the precise localization and three-dimensional structure of the inner body have remained elusive. Here, we applied the high-dose cryo-electron microscopy (“bubblegram”) approach to localize the inner body within the capsid of phiKZ-like jumbo phage phiK601. The inner body was resolved as a cylindrical structure approximately 22 nm in diameter, tilted by ~ 20° relative to the tail axis and positioned asymmetrically, likely contacting the portal vertex and the opposing capsid edge. It is surrounded by 17 concentric layers of packaged DNA and exhibits positional flexibility within the capsid.

Efficient sampling of low-energy conformational states remains a central challenge in protein folding simulations. In this study, we investigated the performance and limitations of standard Monte Carlo (MC) and parallel tempering Monte Carlo (PTMC) methods in exploring protein energy landscapes using two representative systems: the small peptide Met-enkephalin and the FBP28 WW-Domain protein. For Met-enkephalin, standard MC simulations achieved equilibrium at higher temperatures but showed insufficient sampling in low-energy regions. In contrast, PTMC significantly improved sampling efficiency across a wide temperature range, particularly in the low- energy regime, enabling robust exploration of the energy landscape. PTMC successfully identified the global minimum structure of Met-enkephalin with an energy of -11.23 kcal/mol and a root-mean- square deviation of 1.18 Å from the reference structure. The applicability of PTMC to larger proteins was further examined through extensive simulations of the WW-Domain, revealing pronounced energy fluctuations and folding–unfolding transitions at higher temperatures, while sampling at lower temperatures remained limited. The number of visits per unit energy range for two consecutive temperatures have also monitored to check the exchange rate and conformational sampling of the entire landscape. The average energy and specific heat with respect to temperature have also calculated to check the folding unfolding transition of WW-Domain. These results demonstrated the superiority of PTMC over standard MC for protein folding studies; however, they also highlighted challenges associated with larger protein systems.

Protein secondary structure has long been viewed as a local property encoded in the amino acid sequence. While classical definitions of α-helices and β-sheets and established experimental and computational methods have provided a solid framework for structural analysis, growing evidence indicates that secondary structure is strongly influenced by intra- and intermolecular interactions. Such context dependence poses significant challenges for conventional sequence-based prediction approaches. In this review, I discuss the limitations of traditional secondary-structure prediction methods and highlight recent advances enabled by deep learning–based structure prediction, with a particular focus on AlphaFold. Using oligomeric proteins and polytopic membrane proteins as representative examples, I illustrate how interaction-dependent secondary structures can be revealed by integrating computational predictions with experimental data. These observations emphasize the importance of considering molecular interaction contexts to achieve a comprehensive understanding of protein secondary structure and its functional roles.

To reveal functional properties of biological complexes using single-particle cryo electron microscopy (cryo-EM), it is important to obtain maps without distortions. 3D reconstructions can be obtained from back-projection of the individual 2D projections of the particles, a concept that assumes equally distributed orientations. However, most biological samples exhibit preferential views due to preferred particle orientations, e.g., at the air–water interface or on a carbon surface of a cryo-EM grid. Here, we present a method based on Voronoi diagrams that takes into account the Euler angle distribution of particles through individual image weighting, thus reducing 3D reconstruction artefacts emanating from preferential particle orientations. Voronoi diagrams are built from the Euler angle values plotted on a sphere in order to graphically display particle orientations, and the assignment of the projection weights is then calculated, being proportional to the individual Voronoi polygon areas. We implemented this real-space approach into a 3D reconstruction routine that allows performing automated weighting. The method was validated using a back-projection algorithm on synthetic model data and on real experimental data with different point group symmetries and in the presence or absence of preferential views, revealing a particular benefit for asymmetric objects.